Comparison engines pack the power of aggregators--but watch that ROI
By Mary Wagner
If Internet retailing were a board game, the goal would be labeled "ROI," and comparison shopping engines would be one way merchants could try to get there. Like any route to a game's end, it's a path with potential opportunity: It can attract highly targeted shoppers that produce what some say are higher conversion rates. And it's got obstacles: the difficulty for some retailers of wrangling product feeds into the engines and the fact that costs per click, as in other web-based ad programs, are going up.
Among online retailers, comparison shopping engines have proponents like Jacob Hawkins, vice president of online marketing at Overstock.com, who says the price comparison feature is a good fit for Overstock's value-pricing business model. "Some merchants don't want to advertise on comparison shopping engines because they are not trying to compete on price. But we're king of the CSEs," he says. "We love it when people go to these engines to compare prices because it increases our value proposition to customers." The closer a shopper is to making a buying decision, the more targeted the traffic, and the better the chance that the shopper will buy something once he's reached a retailer's site through comparison shopping engines, Hawkins figures.
But as an online advertising venue, comparison shopping engines are losing favor over rising click costs among some earlier adherents. Jay Katz, president of CosmeticMall.com, says comparison sites used to drive a significant portion of CosmeticMall's business, but it's recently scaled back on the product listings it sends to such sites due to higher click fees and diminished ROI.
Selectivity
Instead of sending comparison sites its full 19,000-product catalog, CosmeticMall now limits what it feeds to those sites to about 400 products whose price is high enough to produce an acceptable margin on higher cost-per-click rates. "Even at that, the program is slightly ROI negative and we have to write it off as an advertising expenditure," Katz says.
Comparison shopping engines have been a big win for some retailers and less so for others depending on a host of variables such as product type, marketing strategy, and resources, including IT resources at the merchant end to handle product feeds. But as the growth of search marketing revenue flattens, the operators of ad-driven online marketplaces are looking for new places to grow business. Consider eBay Inc.'s acquisition of Shopping.com for $620 million and media conglomerate E. W. Scripps Co.'s acquisition of Shopzilla for $525 million, both within the past six months. Online marketers, meanwhile, are always looking for sales, from as many sources as possible. Those factors could combine to position the comparison shopping engine industry, now a fraction of the size of the online search market, for significant growth if it can find its way around some barriers. If the industry soars, it will be because it's producing results for marketers--but the extent to which any marketer benefits will be a function of paying very close attention to its comparison shopping program.
Shoppers who go online to look and eventually buy span a range of purchase-readiness, from the casual browser to the focused seeker of a specific product, perhaps even on specific terms. It's that far end of the spectrum that constitutes the sweet spot for comparison shopping engines, and it's the engines' ability to deliver those highly focused shoppers to marketers that keeps marketers willing to pay for that exposure.
Depending on spiders
Comparison shopping engines, whether dedicated destinations such as Shopping.com or attached to portals such as Yahoo, are a vertical search tool specific to retail. They differ from general search engines because they eliminate any search results that aren't about products or merchants. They also provide content and features not attached to general search; for instance, the ability to see prices, product features and customer-generated product reviews and merchant ratings, in an easily comparable format.
According to August figures from comScore Networks Inc., Shopping.com gets the most traffic, with 21.6 million visitors; followed by Shopzilla at 16.7 million; Yahoo Shopping's comparison section at 13.3 million, and NexTag at 10.1 million. ComScore also tracks at least six other comparison sites. Based on their tenure in the marketplace, how they have marketed themselves to shoppers and merchants, and where they've invested resources, different engines have acquired reputations for strength in different product areas.
Like paid search, they charge retailers a fee whenever a visitor clicks on the retailer's listing in search results. Visitors have the freedom to search and sort results by price, product specifications and other features, depending on the functionality offered by the particular comparison site. The default order in which each site presents results, however, is based on a combination of relevancy and in some cases bid price that varies from engine to engine.
To populate their shopping-only indices, comparison shopping sites depend in part on their spiders to grab information from the web, but they also rely in large measure on direct feeds from merchant's product databases. At Shopping.com, for example, about 75% of the merchant offers in the index come from product data feeds, according to Iggy Fanlo, president of worldwide field operations. The feeds get the products into the indexes faster than spiders do, which means they give the engine and the shopper using it a more current view of changing inventory status at merchants. In fact, Shopzilla re-builds its index to add data as often as six times a day, with plans to increase that to 10 soon, according to Farhad Mohit, co-founder and chief products officer of Shopzilla.
The advantage
If a marketer's products can be found quickly and easily through general search--where most Internet shoppers start anyway--not to mention affiliates and banners, what do marketers get by being on comparison shopping engines that they don't get by being at the other ad venues? For one thing, though price comparisons are a popular draw at the shopping engines, the shoppers using the engines are among the web's most well-heeled. ComScore's August figures show that at three of the four-most visited shopping engines, the largest segment of the audience had annual household incomes exceeding $100,000.
Comparison shopping engine operators say the engines also help marketers attack the problem of customer acquisition from more than one angle. Search marketing is a volume play that produces good conversions because of the declaration of intent. Comparison shopping leverages the notion of shopper intent further. As a consequence, it drives less volume because it pulls in only focused shoppers already well down the path to purchase but, theoretically, those shoppers produce better conversion rates.
How much better? Brian Smith, a comparison shopping engine industry consultant and blogger who also manages online marketing for a b2c site that sells home security equipment, says he has seen ROI for the store, PersonalProtectionStore.com, go two to five times higher on the advertising it does on comparison sites than on straight pay-per-click advertising it does on general search engines Yahoo and Google. Smith also notes that those results aren't consistent across all seven comparison sites on which the store advertises, and that he watches results very carefully, cutting back on bids and pulling out of engines when results aren't justifying what he's spending.
Fanlo says that across a variety of marketers using Shopping.com, he has seen conversion rates 50% to 200% higher than on general sponsored search. But Fanlo grants that not all marketers that use comparison shopping engines will see that kind of spread. "The softer the product category, the closer the ROI will be between the two," he says. "If you are selling something that is very brand-focused, keyword engines might do just as well. If you are selling something that is very specification-focused, like a digital camera, the ROI is going to be better on a comparison engine."
More information
Better ROI is tied to the fact that the format of a comparison shopping engine puts more of the information needed to make a complex purchase--product specs, brand, price, shipping details, and product and merchant reviews--in front of the shopper before the shopper ever chooses which retailer's offer to click on in the results. By contrast, shoppers who click on a keyword listing in sponsored search results don't get most of that information in the search listing and arrive at the site less qualified and less ready to buy--but the ad sponsor must still pay for that click.
Shopping in softer product categories isn't as driven by the comparison of multiple attributes on comparison engines because such products are attached to less structured data; that is, categories of data such as product specs that are standard across all products in the category. Cameras, for instance, are associated with a lot of structured data in comparison engines; something like a down comforter, less so.
But even where a product has fewer attributes that can be compared against the competition's in a grid, comparison shopping engines still bring something to the table that general search doesn't: customer reviews of merchants and products, a point underscored by Mohit of Shopzilla, which launched initially as the business ratings service, BizRate. "If e-commerce is going to thrive on the Internet, you need an intermediary of trust," he says. "Just having a series of links may be a good way to arrange information, but it's not going to be good enough to help people make a decision on where to buy."
Reviews can affect click through dramatically, says Stephen Imbler, CFO of comparison shopping engine NexTag. In buying a camera for his own use on NexTag recently, Imbler first sorted for the model he wanted by price, eliminated the three top-listed (lowest-priced) vendors because they weren't highly rated or didn't have enough reviews yet, and zeroed in on the first in his sort that displayed a four-star rating. He then checked how that vendor had been rated within the past 90 days and then the past 30 days, and finally read some of the customer reviews to see what customers did or didn't like about the merchant he was considering before making a decision.
Weighing the pros and cons
"Can you deal with the things people don't like?" he says. "Maybe the review says the vendor ships slowly. If you don't need it next week, and you'd like to save a couple of hundred dollars, maybe that's okay. Maybe the review says the vendor will try to upsell you when you order. If you're up for that, that's fine. If not, you pick another vendor that doesn't upsell." The point for retailers is that success on comparison shopping engines is about more than lowest price, he says. "The reviews are very important," he says. "Just having the lowest price doesn't mean the retailer is going to have the best-performing offer."
While some merchants attain increased conversions and higher ROI on the traffic delivered by comparison shopping engines, and industry leaders like eBay and media conglomerate Scripps have bet on the future of comparison shopping with big-money acquisitions, some marketers in the trenches are grappling with issues that could hinder that growth unless addressed, and they go beyond the rising cost of clicks that affects many CPC programs.
Hawkins observes one limiting glitch in the system: the absence in some cases of a universal identifier on products that ensures that the products the engines pull from an index for comparison from different merchants are, in fact, the same product, and that shoppers are seeing a true apples-to-apples view. Products in some categories have an associated UPC or model number; books might also have ISBNs. "Go over to sheets, pillows, shirts or pants or a number of other product categories and there are no associated model numbers or UPCs or ISBNs. So when you search for a down comforter, all you see are different pictures of down comforters, with different prices listed for different merchants. You just have to look at the product and pick one that looks good," he says.
A related issue is the lack of structured data in some product categories, one reason that categories that have structured data, such as consumer electronics, have lent themselves more readily to comparison listings. "That was the case three years ago, but all of us in the industry are getting more sophisticated at structuring data. We're providing multiple opportunities for comparison in new categories, whether apparel or home and garden. We're finding it's taking a little time to structure data in these other categories, but it's happening," says Rob Solomon, vice president and general manager of Yahoo Shopping.
Monitoring ROI
Yet another limiting factor is the technical requirements of the data feed, a significant issue for smaller merchants without big IT resources. "You talk to ChannelAdvisor, Channel Intelligence, Mercent and Performics, and they are all doing a great job of managing, submitting and optimizing feeds for companies with annual revenues of $10 million to $25 million plus," Smith says. "But they are not dealing with small businesses. For every large Internet Retailer 400 company out there on the shopping comparison engines, there are about 10 or 20 small businesses with no one helping them out."
While the industry works to address those issues, the best advice for merchants looking to profit on comparison marketing is vigilance on monitoring ROI. When determining the likely gain from listing offers from Overstock on any comparison engine, Hawkins looks at some key factors. For one, different engines have different strengths, he notes. "Price Grabber, for example, is good with books, music and videos. NexTag is good at electronics. Shopping.com is good at some of the softer categories, like home and garden. Shopzilla does well with apparel and accessories," he says. The retailer should look at how much volume overall the engine gets, how much traffic the retailer gets in a comparison site's areas of product strength, how many different advertisers are there for that category on that engine, and finally, how well the traffic from that engine converts.
"We actually go out and put products on all the engines, and as we find which ones don't work, we pull them off," Hawkins says. "It's a comparison-engine-by-comparison-engine decision, a category-by-category decision, and even a product-by-product decision."
mary@verticalwebmedia.com
